Albright hereditary osteodystrophy (AHO) is an autosomal dominant disorder characterized by short stature, obesity, subcutaneous ossifications, and brachydactyly. Some family members have these features in association with resistance to multiple hormones which activate Gs-coupled receptors (pseudohypoparathyroidism type Ia, PHP Ia) while others present with the somatic features alone (pseudopseudohypoparathyroidism, PPHP). AHO is caused by heterozygous Gs-alpha null mutations and most affected patients have a 50% deficiency in Gs-alpha subunit function and/or expression in peripheral tissues (both PHP Ia and PPHP). We have shown that Gs-alpha is imprinted in a tissue-specific manner in mice using mice with a genetic knockout of the Gs-alpha gene Gnas. In some hormone target tissues Gs-alpha is expressed primarily from the maternal allele, which explains why maternal inheritance of Gs-alpha mutations results in PHP Ia while paternal inheritance of these same mutations leads to PPHP. More recently GNAS/Gnas has been shown to be a very complicated gene with multiple imprinted gene products generated by several alternative promoters and first exons. NESP55 is a chromogranin-like protein that is maternally expressed while XL-alpha-s is a paternally expressed Gs-alpha isoform with a long amino-terminal extension. Both are primarily expressed in neuroendocrine tissues. We have shown that NESP imprinting is not established until postimplantation development. Just upstream of the XL-alpha-s promoter is the promoter for a paternally-expressed antisense transcript (NESPAS) that traverses the NESP promoter from the opposite direction. We identified another alternative first exon (exon 1A) that generates paternally expressed untranslated mRNAs and that is a maternal germline imprint mark. This region has unique histone methylation patterns on each parental allele. We showed that PHP Ib (parathyroid resistance in the absence of AHO) is virtually always associated with loss of maternal imprinting of exon 1A. A detailed analysis of GNAS imprinting in PHP Ib patients showed that familial cases tend to only have abnormal exon 1A imprinting, while sporadic cases often have additional imprinting defects involving NESP and XLas. In patients with a paternal imprinting pattern of NESP on both alleles (both promoters methylated) loss of NESP expression results in no further phenotype. In some patients, the imprinting of the XL-alpha-s promoter and its first exon is discordant. We have examined Gnas methylation in mice which do not establish maternal germline imprints (dnmt3L-/-) and show that the whole Gnas locus develops a paternal methylation pattern on both alleles, indicating that imprinting of the whole locus depends on maternal germline imprints. We have made exon 1A knockout mice, which show no effect on imprinting of Nesp or XL-alpha-s, suggesting the presence of two independent imprinting domains, both with maternal germline imprints. Preliminary evidence suggests that deletion of exon 1A on the paternal allele may relieve the paternal imprinting of Gs-alpha in renal proximal tubules. We recently showed that Gs-alpha is imprinted in human thyroid and this explains the presence of mild thyrotropin resistance in PHP Ia and Ib. We originally showed that mice that inherit a Gnas exon 2 insertione paternally are leaner than normal and are hypermetabolic and hyperactive, while mice which inherit the knockout maternally become obese and are hypometabolic and hypoactive. Detailed metabolic studies have shown the paternal exon 2 knockout mice to have increased adiponectin and resistin expression in adipose tissue, increased whole body lipid metabolism, and increased insulin sensitivity in adipose tissue, muscle, and liver. We have created a new mouse line with "flox" sites around Gs-alpha exon 1, and with this have made mice with Gs-alpha specific deficiency. Mice with the knockout on the paternal allele have an opposite metabolic phenotype to those with the exon 2 insertion (obesity, insulin resistance, poor lipid clearance). We believe that the paternal exon 2 phenotype is caused by loss of XL-alpha-s and that this Gs-alpha isoform may play a role in regulating sympathetic activity in mice such that loss of this protein leads to a sympathetic hyperactivity. We have also generated tissue-specific Gs-alpha knockouts in various metabolically active tissues and bone and are presently characterizing the phenotypes of these new mouse models.